4 ms·
I would be very interested in a set of benchmarks and intermediate goals for safe AGI (i.e. AGI that is aligned with human interests) rather than just AGI itsel
by dwohnitmok 4y ago
I would be very interested in a set of benchmarks and intermediate goals for safe AGI (i.e. AGI that is aligned with human interests) rather than just AGI itself.
- sdrg822 4y agoWhich humans?
- mitthrowaway2 4y agoThat's an important question but I think even getting an AI to act in the interest of even one human would require solving 99.9% of the technical difficulty.
- tough 4y agoThere won't be any humans left once AGI is real
- mitthrowaway2 4y agoYes, that's pretty much my point.
- lgas 4y agoI volunteer.
- flandry93 4y agoNo such goals possible. See https://mflb.com/ai_alignment_1/index.html https://mflb.com/ai_alignment_1/index.html